Correct extension tube for a 35mm prime?

<p>Hi all,<br>

I've been looking at negative scanning options, and as I have a d7100, I figure I might try rigging a DSLR system. I currently have a 35mm prime lens but I'm trying to figure out what extension tube to use to get a clear 1:1 image for scanning the negative. I wonder if anyone has any thoughts on this? Does the DX system make a difference?<br>

Thanks in advance!<br>

Oli</p>

<p>To photograph at 1:1, you need a distance of 2f (twice the focal length of the lens) between the lens and the film or sensor. With a 35mm lens, this calls for a 35mm extension tube (or slightly smaller, since you can gain a few millimetres extension by focusing the lens).</p>

<p>1:1 is misleading in this context; to copy a 35mm slide you need less magification on DX. Sorry I 'm bad at math and guessing something in the 1:2 range i.e. extension tubes about 17 -21mm should be promising.<br>

If you find a formula somewhere put in the size of DX sensor and 24x 36mm to figure out your reproduction ratio and go for bellows draw from there. Get up to 2mm less than needed.</p>

<p>You could use a 16mm extension tube between camera and lens, and an ES-1 Nikon slide adapter in front. The ES-1 threads into a 52mm filter ring, and adjusts from 60 to 74 mm with a sliding tube. With a DX sensor, you are looking for 0.66 magnification, not 1.00. The working distance will be about 50mm, so you'll just make it without cropping too much of the slide.</p>

<p>A more workable combination is a 60/2.8 (AF) or 55/2.8 (MF) Micro-Nikkor lens with a PK-13 (27 mm) extension tube. The ES-1 will work directly on the lens (52mm filter thread) for a full-frame camera, but you will need a 20 mm extension between the lens and the ES-1 for an APS-C sensor. The extension is a simple tube, threaded 52mm female at one end and male at the other. Several companies make individual rings of this sort, found on Amazon, including a Nikon K extension tube set for a pre-AI lens. (don't even think about mounting it directly on the camera. It will damage the aperture coupling mechanism).</p>

<p>Edward is correct, but I believe you need a number.</p>

<p>Focal distances are calculated thusly: 1/Fo + 1/Fi = 1/F, where Fo is the focal distance to the object (your slide), Fi is the focal distance of the image, and F is the focal length of the lens. So 1/35 = 1/Fo + 1/Fi. We also know that Fo/Fi = 1.5 (the crop factor). Substituting and solving, I get Fi = 87.5mm. If your lens focuses to infinity at Fi=35, then with your lens focused to the infinity mark, you will need 87.5 - 35 = 52.5mm of extension tube. And then the lens extends via the focusing mechanism, so some of that 52.5mm can be from the focusing mechanism itself. If the lens focuses as close as maybe 0.3 m (300mm) then it can extend to 1/(1/35-1/300) = 39.6mm, meaning the barrel extends 4.6mm. So if you put a 50mm extension tube on the lens, you should get a 1.5:1 magnification ratio somewhere in the middle of the focusing range of the lens.</p>

<p>Gah! Always check your work! (It would make no sense for Fi to be greater than 2x F! In my haste, I neglected to divide by 1.5.)</p>

<p>So substituting and solving, you get a focal distance of 58.3mm, meaning you need 23.3mm of extension. You can probably get that from a 20mm extension tube and 3.3mm of extension from the focusing mechanism. And thus Jochen's intuitions are pretty darned good!</p>

<p>BTW, the correct focal distance setting on the lens with a 20mm extension tube would be 0.4 m. </p>

<p>G'nite!</p>

The distances as marked on the lens are the distances from sensor or film to the plane of focus., i.e. Fo + Fi + Di, where Di is the distance between the two principle planes of the lens (the things Fo and Fi are measured from). Di (or HH') does not figure in the formula, because the formula assumes a single, infinitely thin lens, and not a complex optical thing consisting of a number of not infinitely thin lenses.<br>That distance Di is something very few lens makers publish. And it changes in many lenses with focusing.<br><br>That distance Di changes with focusing when lenses change their focal length to focus. Changing focal length F also messes with the lens maker's formula that is used above (which takes a known focal length F as fixed anchor for the rest), so we either need to know the focal length at all lens distance settings, or restrict the lens to whatever setting at which we know the exact focal length (usually infinity).<br>That exact focal length is also not the one marked on the lens. That one is a nominal value, rounded to the nearest traditional value, and the true value may differ from that by as much as 10% either way.<br><br>In short: don't use such calculations except as a coarse indication. The precision suggested by a fractional mm in the result is misleading. The result is more than likely off by even more than one whole mm.

<p>Thank you to everyone for their thoughts! It seems a cleaner lens would be the 50mm (I had debated that lens a while ago when getting the 35!) Slightly off topic, but Edward mentioned the Nikon slide adaptor but it only seems to do 135 frames. I was hoping to try and digitise 120 as well. Has anyone had any luck with similar adaptors working for 120?<br>

Thanks again.</p>

<p>Novoflex makes a slide copy frame which can be used with film up to 7x6 cm. It is used in conjunction with a Novoflex Castel-Q focusing rack. (http://www.adorama.com/nvsca.html)</p>

<p>A problem I see with this arrangement is that extraneous light between the lens and slide is not excluded. This could lead to lower contrast. A possible solution would be to use a bellows type sunshade. Lee makes such a hood for use with their filters. It might not be too hard to make an adapter to fit that hood which would hold a slide or film strip.<br>

http://www.adorama.com/lewalh.html?cvosrc=ppc.google._cat:lenses&matchtype=b&adposition=1t3&creative=52016584065&content=&cvo_search=1&gclid=COyHydjZgscCFVE7gQodLUkH1A</p>

<p>What I've done for experimental purposes is to place the film, in a paper mask, on a light box with the camera held on a tripod pointing downward. I used a compendium hood from an Hasselblad to block light, not attached but sitting on the light table, shooting through the open end. A paper tube, like a 4" core from a minilab, miight work too. Unless you have a rigid fixture, you spend a lot of time fiddling with the film and centering it.</p>

<p>The larger the film, the more effective is a flatbed scanner with a backlight for scanning transparencies. I bought an Epson 500 (550?) for scanning family photos, scrapbooks and documents, which has a transparency adapter. It cost all of $200. Even 2000 ppi will give you a 4200 x 4200 pixel scan. In comparison, a Nikon LS-8000/9000 gives you an 8500 x 8500 scan which is grain-sharp.</p>